For those who might care, The objective here is see how much performance can be wrung from a tube that costs as little as $1.
The other objective here is to see if I can power the tube with a DC-DC converter run by a wall-wart and get decent performance - so far, so good.
This filament voltage of the tube was deliberately chosen to be driven directly from a 12V power supply.
The tube in question is one of several dual dissimilar triodes designed for vertical oscillator./amplifier service that pair a high-mu triode with a linear, beefy low-mu companion. Tubes of particular interest are the 6DR7, the 6CY7, and the 6GF7. There are octal versions of these tubes, but I chose 9-pin variants for compactness. The Noval and Novar tubes also have variants with higher filament voltages for series-string use that can be driven directly from a 12V or 15V wall-wart.
Current source loading the input stage and selecting LED biasing for the output stage provides some excess gain that can be fed back to stiffen the transformer at the low end for better bass response. This could also be done by using an even stiffer semiconductor voltage source in the cathode of the output stage, or just the time-honored cathode bias resistor with heavy bypass capacitor. I may try all these methods to see if there is measurable or audible difference.
Obviously, the "Rolls-Royce" comment set at least one person's teeth on edge - too bad. I'm not shooting for 0.00001% distortion with 50 MHz bandwidth, just a simple, listenable headphone amp. I've already achieved that - I now want to make it as good as can be given the obvious design constraints.
The other objective here is to see if I can power the tube with a DC-DC converter run by a wall-wart and get decent performance - so far, so good.
This filament voltage of the tube was deliberately chosen to be driven directly from a 12V power supply.
The tube in question is one of several dual dissimilar triodes designed for vertical oscillator./amplifier service that pair a high-mu triode with a linear, beefy low-mu companion. Tubes of particular interest are the 6DR7, the 6CY7, and the 6GF7. There are octal versions of these tubes, but I chose 9-pin variants for compactness. The Noval and Novar tubes also have variants with higher filament voltages for series-string use that can be driven directly from a 12V or 15V wall-wart.
Current source loading the input stage and selecting LED biasing for the output stage provides some excess gain that can be fed back to stiffen the transformer at the low end for better bass response. This could also be done by using an even stiffer semiconductor voltage source in the cathode of the output stage, or just the time-honored cathode bias resistor with heavy bypass capacitor. I may try all these methods to see if there is measurable or audible difference.
Obviously, the "Rolls-Royce" comment set at least one person's teeth on edge - too bad. I'm not shooting for 0.00001% distortion with 50 MHz bandwidth, just a simple, listenable headphone amp. I've already achieved that - I now want to make it as good as can be given the obvious design constraints.
I may buy a Magni 3 HP amp just to have a low-priced reference point - it's certainly cheap enough. Given its description and seeing that it takes 14VAC for power, I can almost guess what's inside. If I end up getting one, I'll certainly be popping the hood. It boasts 2W max output, though given the cost constraints I suspect that that's AB power, not Class A, though it probably operates in Class A at lower powers.
It's my intent to focus on Class A VT designs here, though those will certainly be more expensive than the little Schitt HP amp, simply due to the cost of output iron. I may use what I learn from the Magni in another thread, though.
It's my intent to focus on Class A VT designs here, though those will certainly be more expensive than the little Schitt HP amp, simply due to the cost of output iron. I may use what I learn from the Magni in another thread, though.
There is a tool for that, better than posting bitter words.
Attachments
Attached are the updated/corrected schematics of the two versions of this amp - a parafeed option and a plain vanilla 2-stage single-ended amp. 
Both designs employ some feedback around the output transformer to stiffen up the bass response. I'll looking for better output iron, as the Transcendar transformers are definitely a budget option, as shown in the earlier gain-phase plots.
Having said that, the simple single-ended amp enjoyed a good reception at the latest Burning Amp. It was checked out by a number of listeners, some of whom brought their own favorite headphones. Those who didn't listened to the amp with AKG240 Studio headphones.
The parafeed design is doubly challenging, as it needs both good output iron and a good coupling cap. I'm currently using Wima MKP4 caps for output coupling on the parafeed design. Another option might be some Russian paper-oil "tub" capacitors, though opinions are definitely divided on the merits of the Russian paper-oil caps.
Both designs employ some feedback around the output transformer to stiffen up the bass response. I'll looking for better output iron, as the Transcendar transformers are definitely a budget option, as shown in the earlier gain-phase plots.
Having said that, the simple single-ended amp enjoyed a good reception at the latest Burning Amp. It was checked out by a number of listeners, some of whom brought their own favorite headphones. Those who didn't listened to the amp with AKG240 Studio headphones.
The parafeed design is doubly challenging, as it needs both good output iron and a good coupling cap. I'm currently using Wima MKP4 caps for output coupling on the parafeed design. Another option might be some Russian paper-oil "tub" capacitors, though opinions are definitely divided on the merits of the Russian paper-oil caps.
Attachments
I made an inquiry to Cinemag about their offerings for output transformers. What they came up with was the CM-27101, with a 7:1 turns ratio, which is not quite optimum for my purposes (I'd rather have something in the range of 8:1 to 10:1). Their standard offering appears to be gapless, which would suit the parafeed design. What I find troubling from the data sheet is the massive phase shift that shows up right around 100kHz, which would put this transformer in the same league as the Transcendars.
I think I'll be talking to Electra-Print nest, though I'll have to be digging much deeper in my pockets for one of Jack's designs. I'd really like a transformer with more bandwidth, as I'm applying some global feedback.
I think I'll be talking to Electra-Print nest, though I'll have to be digging much deeper in my pockets for one of Jack's designs. I'd really like a transformer with more bandwidth, as I'm applying some global feedback.
PRR, thank you so much for this forum tip and "tool".
I just added you and pipe wrenchie to my IGNORE LIST!
Don't you just love it when you learn something new and there's no need in posting "bitter words" in future replies? I certainly do.
Thank you again PRR and Merry Christmas...LOL.
PRR, ... and pipe wrenchie to my IGNORE LIST!... and Merry Christmas...LOL.
Merry Christmas to you too!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.